Fault Diagnosis With Robustness and Lightweight Synergy Under Noisy Environment

Over the years, deep learning-based fault diagnosis approaches have made remarkable progress. However, the diagnosis accuracy has been on a steady decline because of noise interference in industrial applications. In addition, the accuracy improvement is often accompanied by an increase in computational cost, which makes it difficult to achieve engineering-embedded deployment. To find the balance between robustness and complexity, a new intelligent fault diagnosis framework named lightweight multiscale dilated causal convolutional attention network (Li-MDCCAN) is proposed. Based on dual fusion mechanisms, the proposed network realizes high robustness by adopting multiscale strategy and attention modules, which explores multiscale and multidimensional features along three parallel paths. Furthermore, lightweight convolutional blocks are introduced to reduce the network’s complexity. The performance of Li-MDCCAN is verified with six different datasets and compared with those of four advanced fault diagnosis networks. The experimental results indicate the effectiveness and superiority of Li-MDCCAN in increasing robustness and decreasing complexity. Moreover, prior sensor placement is discussed to guide applications under limited engineering conditions.